THE DAWN OF AUTONOMOUS VESSELS: STEERING TOWARDS STRICT LIABILITY MODEL

By Pushpendra, a fourth year student at Hidayatullah National Law University.

---

Introduction

The maritime industry stands on the precipice of a technological revolution with the advent of autonomous ships. As artificial intelligence and advanced navigation systems evolve, the proposition that autonomous ships can be economically beneficial is becoming a focal point in maritime discussions. Traditional maritime law, which has been developed with human-operated ships in mind, now faces the challenge of accommodating these new, unmanned vessels. The maritime industry, long regarded as the backbone of global trade, is on the brink of a profound transformation. These self-navigating vessels, powered by artificial intelligence, advanced sensors, and satellite communications, promise to revolutionize global shipping by enhancing efficiency, cutting costs, and minimizing human error.

However, the integration of autonomous ships into the maritime industry presents both significant opportunities and complex challenges. As technology continues to advance, it is imperative that maritime law evolves in tandem to address these new realities. In this backdrop, the article strives to examine the legal vacuum created by the rise of autonomous maritime vessels, specifically analyzing the inadequacies of current liability frameworks. It aims to advocate for updated international codes to ensure safe integration of unmanned ships into global trade while proposing a shift towards strict liability approach.

Definition and Governing Framework

Autonomous ships, also known as unmanned or crewless ships, are vessels that can navigate and operate without a human crew on board. The term encompasses various levels of autonomy, from remote-controlled ships to fully autonomous vessels which operate independently without human intervention. Technological advancements in AI, navigation systems, and communication technologies have made the concept of autonomous ships feasible. These ships are equipped with sophisticated software and hardware that allow for real-time decision-making, route planning, and collision avoidance. The International Maritime Organization (‘IMO’) has adopted the term Maritime Autonomous Surface Ships (‘MASS’) for these vessels. A MASS is defined as a vessel capable of being operated without a human on board in charge and which has alternative control arrangements available. The classification typically spans from manned levels through operated, directed, delegated, and monitored, to fully autonomous operation.

The IMO has recognized the need to address the legal implications of autonomous ships. The Maritime Safety Committee is actively working on creating a regulatory framework to ensure the safe operation of these vessels. This includes revising existing regulations and developing new ones tailored to autonomous operations. The scoping exercise for MASS aims to determine how safe, secure, and environmentally sound operations might be addressed in existing instruments. This involves a comprehensive review of conventions such as the International Convention for the Safety of Life at Sea, the International Convention for the Prevention of Pollution from Ships, and numerous others. Currently, several key international conventions govern maritime operations, including the UNCLOS, the Collision Regulations, and the Hague-Visby Rules. Each of these conventions will need to be re-evaluated and potentially amended to accommodate the unique aspects of autonomous ships.

The Liability Conundrum in a Crewless Era

One of the primary challenges is defining what constitutes a ship under international and national laws. Traditional definitions are based on the presence of a crew and human operators. Autonomous ships, which lack these elements, challenge these definitions. It must be determined whether these vessels can be classified as ships and if specific regulations apply to them. Under the UNCLOS, for instance, a ship must be registered to a flag state and adhere to that state's regulations, yet ensuring compliance without human oversight presents a unique challenge. Determining liability in the event of an accident or incident involving an autonomous ship presents a significant legal void.

Traditional maritime law holds the shipowner, master, and crew accountable for the vessel's operation. In the case of autonomous ships, liability might extend to the manufacturers of the AI systems, the operators of shore-based control centers, or even the software developers. The Hague-Visby Rules stipulate the carrier's responsibilities regarding the seaworthiness of the ship, but for autonomous ships, the focus shifts from the crew's competence to the reliability and security of the AI systems. Similarly, safety conventions mandate that ships must be manned by a sufficient number of qualified crew members. For autonomous ships, the challenge lies in demonstrating that AI and automated systems can meet or exceed these safety standards.

Codifying the Future: The Race for Regulatory Standards

In order resolve the various emerging issues associated with autonomous ships, there is a growing consensus on the need for a dedicated Autonomous Ship Code. This code would provide specific regulations for the construction, operation, and liability of autonomous ships, ensuring clarity and consistency in governance. The regulatory scoping exercise is crucial for identifying gaps in current regulations and suggesting amendments. This initiative aims to determine how safe, secure, and environmentally sound operations might be addressed globally.

Different countries are approaching the regulation of autonomous ships in various ways. Norway has established a test area for autonomous ships and is actively developing regulations. Their pragmatic approach includes real-world testing and the development of specific guidelines to integrate autonomous vessels into existing maritime traffic safely. The United States Coast Guard is working on guidelines and policies to ensure that autonomous ships can be safely and efficiently integrated into U.S. waters. This includes focusing on cybersecurity, communication, and navigation standards. In Turkey, the Commercial Code defines a ship based on criteria including size and purpose. It is currently being evaluated whether autonomous ships meet these criteria and how registry requirements will need to be updated.

From Theory to High Seas: Autonomous Vessels in Action

Several autonomous ship projects are currently in development or operation, providing valuable insights into practical challenges. The Svitzer Hermod is a remote-controlled tugboat operating in Copenhagen that demonstrates the feasibility of remote-controlled operations in busy port environments. The Yara Birkeland is known as the world's first fully electric and autonomous container ship designed to reduce emissions and improve safety, serving as a model for sustainable autonomous shipping. The ReVolt is another concept for an autonomous, zero-emission vessel designed for short-sea shipping, highlighting the potential for efficient operation without human intervention.

The well-known Evergreen case involving the collision between the Ever Smart and the Alexandra I serves as a critical case study to highlight the challenges algorithm-based navigational systems will face. In this collision, the outbound Ever Smart collided with the inbound Alexandra I outside the Jebel Ali port channel. If we apply a thought experiment where both vessels were autonomous, the outcome might have been different. The failure of the Ever Smart to keep a good visual lookout could have been resolved with thermal and infrared cameras which can identify objects the human eye cannot. These cameras would have picked up the heat signature of the Alexandra I significantly earlier than the human master who only saw her three seconds before collision. Furthermore, autonomous vessels using enhanced sensors and predictive control algorithms could track and anticipate vessel movements even without standard signals. This technology radically minimizes the risks generally associated with human operators.

The Way Forward: Towards a Strict Liability Regime

The central challenge in governing these vessels lies in defining the locus of control. The important consideration is not just whether persons are on board but whether people on board or ashore have navigational control. While it seems reasonable that normal regulations should apply if an autonomous vessel encounters a manned vessel, any encounter would likely start in machine mode. The current system, where a human operator intervenes only in emergencies, has drawbacks as last-line operators may struggle to make sense of a largely opaque system during a critical failure. Therefore, until the human-in-the-loop role is substantially refined, the expectation is that the human standard will continue to apply, creating significant legal ambiguity. 

This ambiguity complicates the determination of liability for accidents arising from technological defects. Under the existing framework, liability is apportioned based on fault, but since the design of an autonomous vessel involves multiple parties ranging from software programmers to sensor manufacturers, identifying the responsible party for a defect is tedious and expensive. Consequently, a viable alternative is to hold the shipowner strictly liable for all claims arising from the operation of the autonomous vessel. By putting a fully autonomous vessel to sea, the shipowner assumes the risk of the technology. This strict liability regime would simplify the recovery of losses for claimants and allow shipowners to protect themselves through nuanced insurance policies, all while maintaining a right of recourse against manufacturers for technology failures.